Adult mammals, including man, become permanently paralyzed after spinal cord injury because axons from he cerebral cortex and brainstem (supraspinal axons) and axons from spinal cord above the lesion (propriospinal axons) do not grow through the site of injury. In our first specific aim, we propose to employ retrograde and orthograde tracing techniques to establish whether supraspinal and propriospinal axons grow through a complete transection of the thoracic cord during development. If growth through the lesion occurs, as suggested by pilot data, we will ask whether the critical period for it ends earlier than that already established for growth around a lesion; whether the axons which Cross the lesion are regenerative sprouts of Cut axons, late growing axons, or both; whether axons which cross the lesion find their appropriate pathways and terminal targets; and whether they support locomotor function in the adult animal. Once the critical period for growth across a lesion is established, it should be possible to determine the factors which limit it with age and, perhaps, to recapitulate the potential for developmental plasticity in the injured spinal cord of adult mammals, including man. Most supraspinal neurons die after transection of their axons during early development, presumably limiting he degree to which regeneration of cut axons occurs, but susceptibility to axotomy decreases with age. It is our hypothesis that this phenomenon can be explained by a greater dependency on target-derived trophic actors during development than in the adult animal. In our second specific aim, we will seek to establish whether specific neurotrophins (BDNF and/or NT-3) rescue supraspinal neurons from axotomy-induced cell death during development and, if so, whether that effect decreases with age. North American opossums will be employed for the above experiments because they are born in a very immature state (I 2 days after conception) and most descending spinal axons grow into the spinal cord postnatally in opossums, rather than prenatally as in more commonly used laboratory mammals. It is possible, therefore, to manipulate the spinal cord at very immature stages of development without subjecting she mother to intrauterine surgery.